Centered Capability
Statistical indices quantify manufacturing process performance by comparing actual output spread against customer specification limits. Within quality engineering, the cpk capability metric measures how close a process runs to its specification limits relative to natural variability. An index value above 1.33 indicates that process variation remains well inside target thresholds.
Boundary Deviation
Off-center process distributions reduce the margin between measured peak values and maximum tolerance limits. When calculating the cpk capability metric, engineers evaluate both upper and lower specification limits independently to select the lesser value. Process drift toward either limit lowers the final metric regardless of total variance.
Sample Dispersion
Calculation relies on sample standard deviation alongside the empirical mean collected during controlled production runs. Calculating the cpk capability metric requires a stable process under statistical control without special cause variation. Standard calculations assume a normal distribution of continuous sensor readings across test lots.
Non-normal distributions require mathematical transformation before computing tolerance margins. Systemic bias in calibration equipment skews the estimated mean and artificially depresses the resulting ratio.
Acceptance Margin
Industrial suppliers establish minimum ratio thresholds during initial part qualification protocols. A cpk capability metric exceeding 1.67 satisfies stringent automotive and aerospace quality audits. Lower values mandate process adjustment or full automated inspection.
